RL-MPCA: A Reinforcement Learning Based Multi-Phase Computation Allocation Approach for Recommender Systems.
Jiahong ZhouShunhui MaoGuoliang YangBo TangQianlong XieLebin LinXingxing WangDong WangPublished in: WWW (2023)
Keyphrases
- reinforcement learning
- recommender systems
- function approximation
- collaborative filtering
- state space
- reinforcement learning algorithms
- partially observable domains
- model free
- learning algorithm
- optimal policy
- action space
- temporal difference
- action selection
- control problems
- rl algorithms
- multi agent
- user preferences
- markov decision processes
- user profiling
- learning process
- autonomous learning
- learning agents
- recommendation algorithms
- partially observable
- learning capabilities
- transfer learning
- dynamic programming
- principal component analysis
- resource allocation
- function approximators
- temporal difference learning
- reinforcement learning methods
- learning problems
- markov decision problems
- multi agent reinforcement learning
- actor critic
- policy search
- state and action spaces
- user profiles